The International Energy Agency's “World Energy Outlook” predicts that global primary energy demand will increase by 1.7% per year from 2000 to 2030. It also predicts that 90% of this demand will be met by fossil fuels. Consequently, there will be a 1.8% per year increase in carbon dioxide from 2000 to 2030, reaching 38 billion tonnes in 2030. Cleaner, renewable energy sources, including solar cells, have long been heralded as counters to this increased pollution trend. While advanced silicon based solar cells are now widely commercially available, their uptake has been slow due to high production costs, a lack of robustness and associated visual pollution resulting from the large surface exposure requirements.
Dye Sensitised Solar Cells (DSSC) are an alternative to crystalline solar cells that are cheaper than crystalline solar cells to produce. However, DSSC's are less efficient than crystalline solar cells. Therefore, DSSC's require significant area coverage to be effective power generators.
U.S. Pat. No. 4,927,721 entitled “Photo-Electrochemical Cell”, by M Gratzel et al. discloses a typical DSSC. As illustrated in FIG. 1, the DSSC 10 comprises a first transparent insulating layer 1; a first transparent conductive oxide (TCO) electrode layer 2; a transparent metal oxide layer 3 of titanium dioxide (TiO2); a molecular monolayer of sensitiser (dye) 4; an electrolyte layer 5; a second transparent conductive oxide (TCO) electrode layer 6; and a second transparent insulating layer 7.
A DSSC generates charge by the direct absorption of visible light. Since most metal oxides absorb light predominantly in the ultra-violet region of the electromagnetic spectrum, a sensitiser (dye) 4 is absorbed onto the surface of metal oxide layer 3 to extend the light absorption range of the solar cell into the visible light region.
In order to increase the amount of light that the metal oxide layer 3 and the sensitiser (dye) layer 4 can absorb, at least some portion of the metal oxide layer 3 is made porous, increasing the surface area of the metal oxide layer 3. This increased surface area can support an increased quantity of sensitiser (dye) 4 resulting in increased light absorption and improving the energy conversion efficiency of the DSSC to more than 10%.
An electrochromic display (ECD) is a relatively new electrochemical, bi-stable display. While the application is different to the DSSC, these devices share many physical attributes, illustrated in FIG. 1, exchanging the sensitiser (dye) layer 4 by an electrochromic material layer which undergoes a reversible colour change when an electric current or voltage is applied across the device; being transparent in the oxidised state and coloured in the reduced state.
When a sufficient negative potential is applied to the first transparent conductive oxide (TCO) electrode layer 2, whilst the second transparent conductive electrode oxide (TCO) layer 6 is held at ground potential, electrons are injected into the conduction band of the metal oxide semiconductor layer 3 and reduce the adsorbed molecules (the coloration process). The reverse process occurs when a positive potential is applied to the first transparent conductive oxide (TCO) electrode layer 2 and the molecules become bleached (transparent).
A single electrochromic molecular monolayer on a planar substrate would not absorb sufficient light to provide a strong colour contrast between the bleached and unbleached states. Therefore a highly porous, large surface area, nanocrystalline metal oxide layer 3 is used to promote light absorption in the unbleached state by providing a larger effective surface area for the electrochromophore to bind onto. As light passes through the thick metal oxide layer 3, it crosses several hundreds of monolayers of molecules coloured by the sensitiser (dye) 4, giving strong absorption.
Since the structure of both electrochemical devices is similar, we describe only the method of DSSC manufacture as an example. Equally, this process could be applied with little modification to the ECD manufacture.
In order to manufacture the DSSC 10 illustrated in FIG. 1, a metal oxide layer 3 of several microns thickness is deposited onto the first transparent conductive oxide (TCO) electrode layer 2, using any one of several techniques, such as screen printing, doctor blading, sputtering or spray coating a high viscosity paste. A typical paste consists of water or organic solvent based metal oxide nanoparticle suspensions (5-500 nm diameter), typically titanium dioxide (TiO2), a viscosity modifying binder, such as polyethylene glycol (PEG), and a surfactant, such as Triton-X. Following deposition, the paste is dried to remove the solvent, and then sintered at temperatures up to 450° C. This high temperature process modifies the metal oxide particle size and density, and ensures the removal of the organic binder constituents, such as polyethylene glycol (PEG) to provide a good conductive path throughout and a well defined material porosity. Sintering also provides good electrical contact between the metal oxide particles 3 and the first transparent conductive oxide (TCO) electrode layer 2.
After drying and cooling, the porous metal oxide layer 3 is coated with sensitiser (dye) 4 by immersion in a low concentration (≦1 mM) sensitiser (dye) solution for an extended period, typically 24 hours, to allow absorption of the sensitiser (dye) 4 onto the metal oxide layer 3 through a functional ligand structure, that often comprises a carboxylic acid derivative. Typical solvents used in this process are acetonitrile or ethanol, since aqueous solutions would inhibit the absorption of the sensitiser (dye) 4 onto the surface of the metal oxide layer 3.
The first transparent conductive oxide (TCO) electrode layer 2, having the porous metal oxide layer 3 and sensitiser (dye) layer 4 formed thereon, is then assembled with the second transparent conductive oxide (TCO) electrode layer 6. Both electrode layers 2, 6 are sandwiched together with a perimeter spacer dielectric encapsulant to create an electrode-to-electrode gap of at least 10 μm, before filling with the electrolyte layer 5. The spacer material is most commonly a thermoplastic that provides an encapsulation seal. Once the electrolyte layer 5, which is most commonly an iodide/triiodide salt in organic solvent, is introduced, the DSSC is completed by sealing any remaining aperture with either a thermoplastic gasket, epoxy resin or a UV-curable resin to prevent the ingress of water and hence device degradation.
Most, if not all, of the materials used to fabricate the DSSC can be handled in air and also under atmospheric pressure conditions, removing the necessity for expensive vacuum processes associated with crystalline solar cell fabrication. As a result, a DSSC can be manufactured at a lower cost than a crystalline solar cell.
The ECD fabrication process is very similar to that for the DSSC, with several exceptions. The porous metal oxide layer 3 is often patterned by screen printing to provide a desired electrode image, allowing the device to convey information by colouring or bleaching selected regions. Additionally, the sensitiser (dye) layer 4 is replaced with an absorbed electrochromophore material layer. Furthermore, a permeable diffuse reflector layer, typically large particles of sintered metal oxide, can be positioned between the first and second electrode layers 2, 6 to increase the viewed image contrast.
U.S. Pat. No. 5,830,597, entitled “Method and Equipment for Producing a Photochemical Cell”, by H Hoffmann also discloses a DSSC 100. As illustrated in FIG. 2, the DSSC 100 comprises a first substrate 101 of glass or plastic; a first transparent conductive oxide (TCO) layer 102; a titanium dioxide (TiO2) layer 103, a dye layer 104; an electrolyte layer 105; a second transparent conductive oxide (TCO) layer 106; a second substrate 107 of glass or plastic; and insulating webs 108, 109. The insulating webs 108, 109 are used to form individual cells 110 in the DSSC 100.
An individual cell 110 formed between the insulating web 108 and the insulating web 109 is different from the adjoining individual cell 110 formed between the insulating web 109 and the insulating web 108. This is because the TiO2 layer 103 and the electrolyte layer 105 are interchanged in each adjoining individual cell 110. Thus, the electrical polarity of the adjoining individual cells 110 is opposite. This alternate division of different layers results in the formation of conducting layers 111 from the electrically conductive layers 102 and 106, each conducting layer 111 connecting a positive (negative) pole of one individual cell 110 to the negative (positive) pole of an adjacent individual cell 110. The resultant structure provides a method of increasing the overall DSSC output voltage, without the necessity of incorporating a multi-layered structure.
In order to improve the incident photon to current conversion efficiency and control the stability/reproducibility of the DSSC performance, it is important to precisely control the physical properties of the metal oxide layer, and hence the absorption of the sensitiser (dye) molecule. However, metal oxide layer fabrication using screen-printing often results in a ±5% film thickness variation caused by residual blocked or dirty screen cells, adhesion to the screen during separation from the substrate surface and trapped bubble expansion during drying, caused by the inability to completely outgas a viscous paste. Other methods, such as doctor-blading, also suffer from an inability to provide a well defined thick metal oxide layer without significant spatial deviations. Subsequent porosity and film quality deviations are therefore likely to occur throughout such metal oxide layers, resulting in a degradation of efficiency and image quality for the DSSC and ECD, respectively.
In the case of the ECD, screen-printing demands are further exacerbated by the requirement to create ever finer metal oxide layer features for higher quality images, i.e. increase the dots-per-inch (dpi) for a pixelated display. As the dpi increases, the smallest feature size becomes limited as the screen mesh size approaches the mesh partition width.
As a result, fabrication of an electrochemical device based on a functionally sensitised thick porous metal oxide layer, as for the DSSC and ECD, using the aforementioned fabrication techniques are inappropriate from the view points of device reproducibility and adaptability to large size device production.